Beyond gain of function: strengthening oversight of research with potential pandemic pathogens

ABSTRACT

In 2017, the Department of Health and Human Services adopted a policy, known as the P3CO Framework, to govern proposed research that could enhance the lethality or transmissibility of a potential pandemic pathogen. The prospect of a human-made virus with artificially enhanced lethality and transmissibility has raised serious biosafety and biosecurity concerns. The COVID-19 pandemic has generated new concerns about the risks posed by such research. Even if the origins of the pandemic are presumed or proven to be the result of a natural zoonotic spillover event, the pandemic has placed greater scrutiny on research that could generate pandemic-capable viruses and dramatically illustrated the consequences if such a virus were released from a laboratory. This article assesses the strengths and weaknesses of the P3CO Framework and provides recommendations for strengthening oversight of research with potential pandemic pathogens. The P3CO Framework should be replaced by a national policy that would apply to all relevant research, regardless of the source of funding and be implemented by a new national agency for biorisk management. This new policy would need to be accompanied by a comprehensive analysis of potential pandemic pathogen risks, clear guidance on how to identify research that falls within the scope of the policy, a rigorous process for reviewing the risks and benefits of such research, and criteria for determining the appropriate measures needed to conduct such research safely, securely, and responsibly.

In the early 2000s, scientists and security experts became increasingly concerned about the potential for dual-use research in the life sciences to be misused to cause harm. This concern peaked in 2011–2012 when scientists succeeded in changing a strain of avian influenza that is normally highly pathogenic, H5N1, to be transmissible between mammals. The prospect of a human-made virus with artificially enhanced lethality and transmissibility raised serious biosafety and biosecurity concerns. The escape of such a virus from a laboratory could spark a major outbreak or even a pandemic. Likewise, knowledge about how such a virus was constructed could provide insights to malicious actors into how to replicate the experiment and create their own potential pandemic pathogen.

The H5N1 controversy highlighted the widely divergent views on the benefits and risks of dual-use research held by different stakeholders, including scientists, publishers, biosecurity experts, the national security community, and public health officials. On the one side, proponents of the research focused on the public health benefits of knowing that H5N1 can be transmitted between mammals and which specific mutations can confer this ability to the virus. Opponents of the research highlighted the risks of a laboratory accident and the potential for a nefarious actor such as a terrorist group or rogue scientist to replicate the research and deliberately release the virus.

Between 2012 and 2017, the United States adopted a number of policies designed to provide greater oversight of dual-use research, including so-called ‘gain of function’ research that can result in a pathogen with a new trait, such as enhanced virulence, pathogenicity, transmissibility, or growth in culture. When these characteristics are enhanced, associated biosafety and biosecurity risks are often greater. The centerpiece of U.S. oversight of this type of research is the Framework for Guiding Funding Decisions about Proposed Research Involving Enhanced Potential Pandemic Pathogens issued by the Department of Health and Human Services (HHS) in December 2017.¹ This policy, commonly referred to as the HHS P3CO Framework, provides for oversight of research funded by HHS that is ‘reasonably anticipated’ to enhance the virulence and/or transmissibility of a potential pandemic pathogen (PPP), which is a pathogen capable of ‘wide and uncontrollable spread’ in human populations and able to cause ‘significant morbidity and/or mortality’ in such a population. Under the framework, funding agencies are supposed to refer grant proposals containing experiments that fall under the scope of the policy to a department-wide review committee to determine if the research should be funded and if any additional risk mitigation measures should be implemented. With this new policy in place, HHS resumed funding research on potential pandemic pathogens, pending review under the P3CO Framework. Three such projects, all involving avian influenza, were approved in 2019 although only two were funded.²

The COVID-19 pandemic raised new questions about the efficacy of this oversight in both principle and practice. One of the potential explanations for the emergence of the SARS-CoV-2 virus is that the virus escaped from the Wuhan Institute of Virology, a biosafety level (BSL)-4 laboratory that conducts research on coronaviruses and is in the Chinese city where the virus was first detected. While this lab received funding from the National Institute of Health (NIH) for collaborating with U.S. researchers to study coronaviruses, including experiments that involved the modification of these viruses, these experiments were deemed by NIH not to be covered by the P3CO Framework. Connections between the lab, the pandemic, and the NIH have become heavily politicized. Even if the origins of the pandemic are presumed or proven to be the result of a natural zoonotic spillover event, the pandemic has placed greater scrutiny on research with viruses engineered to be more virulent or transmissible and dramatically illustrated the consequences if such a virus escaped and sparked a pandemic.

This article is a narrative review that provides a brief history of the origin of the HHS P3CO Framework, assesses the strengths and weaknesses of the policy, in terms of both its content and implementation, and provides recommendations for strengthening oversight of research with potential pandemic pathogens. We recommend that the P3CO Framework be replaced by a national policy that would apply to all relevant research, regardless of the source of funding and be implemented by a new national agency for biorisk management that could conduct a comprehensive analysis of potential pandemic pathogen risks, provide clear guidance on how to identify research that falls within the scope of the policy, oversee a rigorous process for reviewing the risks and benefits of such research, and develop criteria for determining the appropriate measures needed to conduct such research safely, securely, and responsibly.

Origin of the HHS P3CO Framework

Oversight of research with potential pandemic pathogens in the United States progressed in fits and starts between 2012 and 2017, largely in reaction to high-profile incidents. The controversy over the experiments that created strains of H5N1 influenza virus that is transmissible in mammals led to the formulation of a U.S. policy in March 2012 on dual-use research of concern (DURC) which required review of any proposed research conducted by a federal agency on 15 pathogens and toxins (Table 1) that is ‘reasonably anticipated’ to produce one of seven experimental outcomes (Table 2).³ The list of pathogens is based on the Tier 1 high-consequence biological threats designated by the Federal Select Agent Program (FSAP). The seven experiments of concern are based on those proposed by an influential 2004 National Academies of Science report, Biotechnology Research in an Age of Terrorism (also known as the Fink report after the chair of the study Gerald R. Fink).⁴ This DURC oversight was extended to Federally funded research institutions in September 2014.⁵

Table 1.

Tier 1 Select Agents subject to DURC Oversight.

Bacillus anthracis

Bacillus cereus Biovar anthracis

Botulinum neurotoxin

Burkholderia mallei

Burkholderia pseudomallei

Ebola virus

Foot-and-mouth disease virus

Francisella tularensis

Marburg virus

Rinderpest virus

Toxin-producing strains of Clostridium botulinum

Variola major virus

Variola minor virus

Yersinia pestis

Table 2.

Dual-use research of concern (DURC).

1. Enhances the harmful consequences of the agent or toxin

2. Disrupts immunity or the effectiveness of an immunization against the agent or toxin without clinical or agricultural justification

3. Confers to the agent or toxin resistance to clinically or agriculturally useful prophylactic or therapeutic interventions against that agent or toxin or facilitates their ability to evade detection methodologies

4. Increases the stability, transmissibility, or the ability to disseminate the agent or toxin

5. Alters the host range or tropism of the agent or toxin

6. Enhances the susceptibility of a host population to the agent or toxin

7. Generates or reconstitutes an eradicated or extinct agent or toxin

In response to biosafety and biosecurity concerns generated by the H5N1 experiments, virologists voluntarily paused this type of research between January 2012 and January 2013.⁶ Since highly pathogenic avian influenza was not listed as a Tier 1 Select Agent, the 2012 DURC policy did not cover research with that virus that might generate more virulent or transmissible strains. Therefore, in February 2013, HHS issued a framework outlining its process and criteria for funding research with mammal-transmissible strains of H5N1 HPAI.⁷ In August 2013, following a large outbreak of H7N9 avian influenza in humans in China, HHS extended this oversight framework to cover similar research with H7N9.⁸

A series of subsequent, unrelated biosafety incidents in 2014 led to a pause in US government funding of so-called ‘gain of function’ research and a multi-year process to devise a system for overseeing such research in the United States. The National Science Advisory Board for Biosecurity (NSABB), a multidisciplinary group of non-governmental experts who advise the NIH on biosecurity issues, was tasked with developing recommendations for conducting gain of function research safely, securely, and responsibly. In support of the NSABB deliberations, the National Academy of Sciences held two symposia,⁹ Gryphon Scientific was commissioned to provide a comprehensive analysis on the risks and benefits of gain of function research,¹⁰ and an expert in bioethics conducted an ethical analysis of the funding and conduct of such research.¹¹ NSABB issued its recommendations in May 2016.¹²

In January 2017, on the basis of the NSABB recommendations, the Office of Science and Technology Policy (OSTP) issued ‘Recommended Policy Guidance for Departmental Development of Review Mechanisms for Potential Pandemic Pathogen Care and Oversight (P3CO)’ which established the minimum standards that Federal agencies would need to adopt to resume funding of research that could possibly enhance the virulence or transmissibility of potential pandemic pathogens.¹³ Since the OSTP guidance is not itself operational, it has to be implemented by an agency through the development of a review mechanism in order for the agency to resume funding research covered by the 2014 funding pause.¹⁴ This OSTP guidance provided the basis for the HHS P3CO Framework that was issued in December 2017.¹⁵

In January 2020, NSABB was tasked with reviewing dual-use research oversight in the United States, including the HHS P3CO Framework.¹⁶ This review was cut short by the onset of the COVID-19 pandemic. In January 2022, the NIH and OSTP launched a new review of dual-use research oversight in the United States, including the HHS P3CO Framework. As part of this evaluation, the NSABB was charged with evaluating the effectiveness of this oversight and determining whether the current approach sufficiently addresses future potential threats in biological research.¹⁷ In January 2023, NSABB unanimously approved a plan to reform the U.S. approach to oversight of dual-use research, including work with enhanced potential pandemic pathogens.¹⁸

Congress has also taken an interest in these issues. In December 2022, President Joe Biden signed legislation requiring OSTP to review and update Federal policy governing research with potential pandemic pathogens by the end of 2023.¹⁹ In January 2023, Republican leaders in the House of Representatives reconstituted a COVID oversight committee to focus on the origins of the pandemic which will include scrutiny of U.S. oversight of ‘gain of function’ research in the United States and China.²⁰ In May 2023, on the eve of launching his 2024 presidential campaign, Governor Ron DeSantis in Florida signed legislation banning research that is ‘reasonably likely’ to create a potential pandemic pathogen with enhanced virulence or transmissibility.²¹

Strengths of the HHS P3CO Framework

There are three positive aspects to the HHS P3CO Framework: it eschews the term ‘gain of function,’ provides a detailed process by which the department-level committee conducts reviews of research proposals under this policy, and it takes a risk-based approach that is based on the attributes of the modified organism.

The first positive aspect of the Framework is that it does not use the term ‘gain of function.’ Carving out ‘gain of function’ as somehow distinct or separate from dual-use research has muddied the debate and continues to cause confusion today.²² The introduction of this term into the discussions on dual-use research in 2011–2012 triggered a long and unproductive debate about how to define this category of research. This phrase is not found in the 2004 Fink report which provided a list of seven experiments of concern that should be subject to oversight.²³ This list provided clearer and more specific guidance for what life sciences research should be considered dual-use compared to the vague and ambiguous phrase ‘gain of function.’ Before the research that created a mammal-transmissible strain of H5N1, gain of function was a scientific term commonly used by biologists to describe a broad array of experiments that resulted in a microbe with new or enhanced attributes or properties. During the debate over the H5N1 research, a subset of gain of function research, called gain of function research of concern (GOFROC), was identified which included enhancing the virulence and/or transmissibility of a pathogen.²⁴ While this new term helpfully delineated the most worrisome types of gain of function research from a biosafety and biosecurity perspective, it never gained traction within the life sciences or biosecurity communities. The P3CO Framework captures the essence of GOFROC by defining increased virulence and transmissibility as being the hallmarks of an enhanced potential pandemic pathogen. While we believe the definition and threshold for research that should be covered by the P3CO Framework should be clarified, this discussion is far more productive than one centered around the ambiguous phrase ‘gain of function.’

Second, the HHS P3CO Framework usefully describes the principles and processes through which the department-wide committee at HHS will review proposed research that involves enhanced potential pandemic pathogens (ePPPs) (see Table 3). The requirement that the department-level review be conducted by a multidisciplinary group of experts is important as is the option to include voting and nonvoting members from within HHS as well as other federal agencies. The inclusion of seven categories of dual-use experiments that the department-level review must also consider usefully expands, beyond increases in virulence and transmissibility, the range of potential risks that will be subject to review by HHS. This list of seven experiments is the same as those included in the 2012 and 2014 U.S. government policies for dual-use research of concern – but those policies are only applied to research with Tier 1 Select Agent pathogens and toxins. Therefore, the HHS P3CO Framework is more in line with the original intent of the Fink Report which envisioned all life sciences research, not just that conducted on a small subset of especially dangerous pathogens, to be subject to oversight for dual-use and biosecurity concerns.²⁵

Table 3.

Criteria for department-level review of Proposed research reasonably anticipated to involve, create, transfer, or use enhanced potential pandemic pathogens.

1) The research has been evaluated by an independent expert review process (whether internal or external) and has been determined to be scientifically sound;

2) The pathogen that is anticipated to be created, transferred, or used by the research must be reasonably judged to be a credible source of a potential future human pandemic;

3) An assessment of the overall potential risks and benefits associated with the research determines that the potential risks as compared to the potential benefits to society are justified;

4) There are no feasible, equally efficacious alternative methods to address the same question in a manner that poses less risk than does the proposed approach;

5) The investigator and the institution where the research would be carried out have the demonstrated capacity and commitment to conduct it safely and securely, and have the ability to respond rapidly, mitigate potential risks and take corrective actions in response to laboratory accidents, lapses in protocol and procedures, and potential security breaches;

6) The research’s results are anticipated to be responsibly communicated, in compliance with applicable laws, regulations, and policies, and any terms and conditions of funding, in order to realize their potential benefit;

7) The research will be supported through funding mechanisms that allow for appropriate management of risks and ongoing Federal and institutional oversight of all aspects of the research throughout the course of the research; and

8) The research is ethically justifiable. Non-maleficence, beneficence, justice, respect for persons, scientific freedom, and responsible stewardship are among the ethical values that should be considered by a multidisciplinary review process in making decisions about whether to fund research involving PPPs.

A third positive aspect of this policy is that it takes a risk-based approach that is not predicated on a list of specific pathogens, but instead focuses on the attributes of the modified organism. While the identity of starting organisms is central to the 2012 and 2014 U.S. government DURC oversight policies, the HHS P3CO Framework emphasizes the importance of a pathogen’s properties once the experiment is over. This more comprehensive approach to dual-use research is a welcome change and should be more broadly reflected in how the United States conducts oversight of dual-use research in general.

Weaknesses in the HHS P3CO Framework

However, the HHS P3CO Framework as currently formulated has three weaknesses and gaps: it applies only to a subset of research conducted with potential pandemic pathogens, it provides an inadequate description of the process and criteria used by the funding agency to determine if a research proposal is covered by the P3CO Framework, and it includes insufficient mechanisms for transparency.

Oversight of potential pandemic pathogens applies only to a subset of Life Sciences Research

The first weakness of the HHS P3CO Framework is that this policy applies only to experiments described in grant proposals funded by HHS, not all research conducted with funding from the department. In addition, the policy does not apply to research that is funded by other Federal agencies or privately funded. As a result, the United States lacks a comprehensive, national policy for providing oversight of research with potential pandemic pathogens with enhanced virulence or transmissibility.

Since OSTP issued its P3CO guidance in January 2017, HHS is the only Federal agency that has developed its own policy for reviewing research with enhanced potential pandemic pathogens and satisfied the requirements of the OSTP guidance to fund such research.²⁶ However, there are several agencies, such as the US Department of Agriculture, the Department of Defense, Department of Homeland Security, US Agency for International Development (USAID), and the Department of Energy that conduct or sponsor research that could involve enhanced potential pandemic pathogens. Indeed, we know of one case where research with an enhanced PPP was proposed to the Defense Advanced Research Projects Agency (DARPA) but it was rejected, due in part to concerns about dual-use issues.²⁷ In addition, USAID’s Discovery & Exploration of Emerging Pathogens – Viral Zoonoses (DEEP VZN) project, announced in 2021, was going to include laboratory characterization of newly discovered viruses, including in vitro analyses of the ability of such viruses to infect humans and possibly in vivo animal transmission studies, to determine the level of zoonotic risk posed by these viruses. ²⁸ USAID stated that DEEP VZN would not conduct ‘gain of function’ research with these novel viruses.²⁹ However, USAID did not define what it considers ‘gain of function’ research nor describe the process by which it will monitor proposed research to determine if the novel viruses are potential pandemic pathogens and if the proposed experiments could be ‘reasonably anticipated’ to lead to enhanced virulence or transmissibility. Since the type of research that DEEP VZN planned on conducting has been associated with ‘gain of function’ experiments with potential pandemic pathogens in the past, its inability to convince officials in the executive and legislative branches that it had an effective biorisk management system in place contributed to the agency’s decision to terminate the program in September 2023.³⁰

The P3CO Framework has been interpreted as applying only to research described in grant proposals that have been submitted to an HHS funding body such as NIH, not to all research conducted by an institution that receives HHS funding. According to HHS, the P3CO Framework ‘is intended to guide HHS funding decisions on individual proposed research that is reasonably anticipated to create, transfer, or use enhanced PPPs.’³¹ When the policy was written in 2017, it was assumed that such research would not be common or widespread and NIH would be the most likely source of funding for such studies. However, due to the pandemic, the creation of chimeric coronaviruses and experiments to determine the virulence, pathogenicity, transmissibility, and susceptibility to vaccines and therapeutics of these viruses have become more common.³²

In 2022, scientists from Boston University published a pre-print article describing how they engineered a strain of an early version of SARS-CoV-2 with the spike protein from the less pathogenic but more transmissible Omicron variant.³³ This new strain, dubbed Omi-S, was less virulent than the original wild-type virus, although significantly more virulent than Omicron, but also more transmissible than the original wild-type strain due to the immune-evasion capabilities conferred by Omicron’s spike protein. Although the pre-print acknowledged that the authors received funding from NIH, Boston University defended its decision not to seek review of this experiment under the P3CO Framework since the NIH grants cited were ‘used to help develop the tools and platforms that were used in this research; they did not fund this research directly.’³⁴ According to NIH, Boston University’s grant proposals to the National Institute for Allergies and Infectious Diseases (NIAID) did not specify they would be conducting such research or mention it in progress reports once the grant had been awarded.³⁵ NIH later ruled that the experiments ‘were funded with private funds and were not subject to NIH review.’³⁶ While a scientists’ research agenda can change after a proposal has been funded, for researchers to be able to use NIH funding to support research with enhanced potential pandemic pathogens as long as they do not report that they will conduct such research when they submit their grant proposals and use private funds to conduct the actual experiment creates a huge loophole in the P3CO Framework. While Boston University may be legally correct in claiming that they complied with all ‘required regulatory obligations and protocols,’ they also violated the spirit and intent of the P3CO Framework by conducting research that could be ‘reasonably anticipated’ to enhance the virulence or transmissibility of a pandemic pathogen without seeking review by HHS.

Another weakness highlighted by the BU episode is that the HHS P3CO Framework is not applicable to research on ePPPs that is privately funded. Given the increasing role of the private sector and philanthropies in funding life sciences research, this is a large and growing loophole.³⁷ The synthesis of horsepox virus by Canadian scientists, with funding from a U.S. biotech company, illustrates how privately funded research can stray into the realm of dual-use research.³⁸ The development of an enhanced strain of Omicron by Boston University also highlights this issue since they claimed that they did not have any obligation to seek review of the experiment since it was conducted with funds from the university, not NIH.³⁹

Since the accidental or deliberate release of an enhanced potential pandemic pathogen could cause a major outbreak, if not a pandemic, there is a compelling public health and safety rationale for regulating such research, regardless of the source of funding. Indeed, in 2016, the NSABB recommended that oversight of research with potential pandemic pathogens be applied to all researchers, regardless of their source of funding.⁴⁰ The 2017 guidance issued by the Office of Science and Technology Policy called for a process to consider the adoption of this NSABB recommendation.⁴¹ The NSABB repeated this recommendation in a report submitted to NIH in 2023.⁴²

Inadequate description of process and criteria for use by funding agency to review research proposals

The P3CO Framework’s description of the process and criteria to be used by an HHS funding agency to determine if a research proposal should be forwarded to department-level review is woefully inadequate. Under the HHS P3CO Framework, the funding agency is responsible for determining if proposed intramural and extramural research is ‘reasonably anticipated to create, transfer, or use’ an enhanced potential pandemic pathogen, which is defined as a pathogen that has been modified to become ‘likely highly transmissible and likely capable of wide and uncontrollable spread in human populations’ and ‘likely highly virulent and likely to cause significant morbidity and/or mortality in humans.’⁴³ However, the framework does not provide any guidance to the funding agency about how to measure how ‘likely’ a pathogen is to be highly transmissible, highly virulent, or highly pathogenic in human populations, or, for that matter, how to characterize the level of transmissibility, virulence, or pathogenicity.

The word ‘likely’ is too strong, as it implies a high probability or a high level of confidence in the estimated probability. ‘Possibly’ would be a better qualifier as it implies some probability but does not set the bar too high. This small change is critical since, absent a large outbreak, laboratory accident, or experimental infection of humans (which raises serious ethical and legal questions), determining ahead of time how transmissible, virulent, or pathogenic a novel virus will be in humans may be impossible.

The HHS P3CO Framework describes the risk of ePPPs largely in terms of transmissibility and virulence, however, these terms are described vaguely, and no suggestions are given on how to assess these traits absent an ongoing outbreak. No definitions of ‘highly transmissible’, ‘capable of wide and uncontrolled spread,’ or ‘significant morbidity’ are given. This lack of precision complicates determining when a pathogen is a PPP. The NSABB recommendations suggest Y. pestis may be a PPP, but this pathogen seems poorly transmissible in modern settings.⁴⁴ Some may consider Ebola and Marburg Hemorrhagic Fever Viruses the canonical example of PPPs due to their ability to create highly concerning outbreaks, but their inability to spread via the respiratory route and the ability to control infection in modern medical settings may cause some to doubt their inclusion in the category of PPPs. Likewise, this lack of clarity hampers the determination of which experimental outcomes would lead to the generation of an ePPP. For example, would increasing the transmissibility of a seasonal influenza strain be sufficient? What about increasing its virulence? If so, by how much?

Finally, there is no guidance for judging when the standard of ‘reasonably anticipated’ is met despite the central role this phrase plays as a threshold that must be crossed for an experiment to be examined under the P3CO Framework. Specifically, the P3CO Framework only applies to ‘proposed research that is reasonably anticipated to create, transfer or use enhanced PPPs’ and the phrase ‘reasonably anticipated’ occurs 11 times in the document. A Government Accountability Office (GAO) report identified the failure to define this key term in either the framework itself or implementation guidance as a key weakness in this policy.⁴⁵ Moreover, the scientific and policy communities may disagree about what a reasonably anticipated outcome is. Biologists are right to emphasize that the phenotypes of transmissibility and pathogenicity are complex, and our state of knowledge is relatively rudimentary. As a result, researchers cannot state with a high degree of confidence that a particular change in a pathogen’s genome would lead to a particular change in phenotype. That is, researchers would not be surprised if they found that any genotypic change in a pathogen did not lead to significant increases in transmissibility or pathogenicity and because of this uncertainty, many experiments do not qualify for being ‘reasonably anticipated’ to generate pathogens with enhanced virulence or transmissibility. At the same time, there are also genes that are closely associated with virulence and transmissibility and so modification of such a gene could be ‘reasonably anticipated’ to lead to enhanced virulence or transmissibility.⁴⁶ For example, despite Australian scientists claiming to be surprised that the insertion of IL-4 into mousepox generated a much more virulent virus, there was ample scientific literature that indicated that this result should not have been unexpected.⁴⁷ Likewise, given the well-known role of the spike protein in Omicron’s immune evasion but uncertain role in pathogenicity, the Boston University researchers should not have been surprised when their recombinant virus manifested the greater transmissibility of Omicron compared to the parental strain.⁴⁸ Anticipating the traits of modified pathogens is especially challenging in experiments expressly designed to characterize novel and existing viruses for their zoonotic potential and to assess the risk of their virulence and transmissibility in human populations or appropriate animal models. This is a reason for, not against, more closely reviewing the benefits and risks of such experiments.

The NSABB’s 2016 recommendations on GOFROC cites example experiments that may be within scope of the framework, but these recommendations are difficult to use when reviewing research.⁴⁹ Some examples also use the term ‘anticipated’ without providing any guidance for the confidence level in experimental outcomes necessary to cross that threshold while other examples are described by their outcome, not at the research proposal phase. Describing experiments by their outcome presumes some unspecified level of anticipation, and therefore could lead to disagreements about the level of uncertainty that prevails. Overall, the lack of detail and ambiguous terminology in the P3CO Framework provides HHS funding agencies with excessive latitude in determining which research proposals fall under the scope of the P3CO Framework – in contrast to the well-delineated process and criteria that are used by department-level group to review the proposals that are referred to them.

Insufficient transparency

The HHS P3CO Framework does not satisfy the guidance from OSTP, which called on agencies to ‘provide transparency to the public regarding funded projects involving the creation, transfer, or use of enhanced [potential pandemic pathogens] … to the maximum extent possible.’⁵⁰ Instead, the HHS P3CO Framework provides only for occasional transparency about the process itself, not the results of that process. According to the HHS P3CO Framework, HHS ‘will periodically ask the National Science Advisory Board for Biosecurity to review the process described herein.’⁵¹ This approach to transparency is not timely nor tied to specific reviews of proposed or funded projects. In addition, there are no guarantees that the NSABB’s review will be made available to the public. Furthermore, since the NSABB is advisory only, it has no authority to require HHS to revise its P3CO Framework, modify its process for implementing the Framework, or revisit or reverse a funding decision. Although the HHS P3CO Framework recognizes the importance of transparency for maintaining public trust in science, it does not go far enough in providing the requisite level of transparency to achieve that objective. The GAO identified several specific areas of the framework that lacked transparency including the composition of the department-level review committee and application of the review criteria by the committee.⁵² Implementation of the NSABB’s 2023 recommendation to improve transparency in the oversight process, including publication of ‘a summary of key determinants that informed ePPP research funding decisions,’ would be a step in the right direction.⁵³

Shortcomings in implementation

Implementation of the HHS P3CO Framework by NIH, the only HHS agency that has funded relevant research, has suffered from two shortcomings: the lack of guidance for outside scientists on how to implement the framework and a lack of transparency on how NIH has implemented the Framework.

Lack of guidance on implementation

HHS funding agencies, namely NIH, have failed to develop, and share with researchers, the public, and policy-makers, clear guidance on how it implements the HHS P3CO Framework. On its website dedicated to ‘Research Involving Enhanced Potential Pandemic Pathogens,’ NIH does not provide any guidance for researchers to help them determine whether their proposed research could be ‘reasonably anticipated’ to involve ePPP work and what the process is for reviewing such proposals under the P3CO Framework.⁵⁴ Nor does such guidance appear to have been provided to NIH staff or outside scientists who participate in the NIH grant proposal review process. NIH officials told the GAO that ‘the criteria for referral in the Framework are well defined and adequate’ and therefore additional guidance was unnecessary. Nonetheless, NIAID felt compelled to develop guidance for its own staff on how to determine whether a research proposal should be referred to departmental review. This guidance reportedly advises NIAD staff ‘to err on the side of inclusion’ when identifying research that may involve enhanced potential pandemic pathogens.⁵⁵ This lack of clear guidance has likely contributed to NIH’s inconsistent approach to reviewing such research, possible failures to refer ePPP work to the HHS-level review committee, and confusion by outside researchers about their reporting obligations.

The only public description of NIH’s process for implementing the HHS P3CO Framework is provided in a July 2021 letter from NIH Director Francis Collins to Senator Charles Grassley:

“At NIH, following the completion of peer review by an SRG [Scientific Review Group composed of outside scientists who perform the first part of NIH’s two-part scientific peer review process], if an SRG identifies research that may create, transfer, or use ePPPs as described above, the Scientific Review Officer records this as an administrative note. NIH Institute or Center program officials with scientific expertise in infectious diseases then review all proposed research found to be scientifically meritorious that is being considered for funding to determine if the research meets the scope of the HHS P3CO Framework. In the case of NIAID, if NIAID program officials determine that the proposed research may meet the scope of the HHS P3CO Framework, the proposed research is further evaluated by a group within NIAID, including members of NIAID leadership, with broader infectious diseases expertise. When evaluating proposed research to determine if it meets the scope of the HHS P3CO Framework, details of the experiment(s) and pathogen(s) are considered in the context of the state of the science in that field. All proposed research determined by this group to fall within the scope of the HHS P3CO Framework is referred by NIAID to HHS.”⁵⁶

This brief description leaves several questions unanswered. What guidance, if any, is provided to the scientists, drawn primarily from academia, who comprise the Scientific Review Groups to enable them to determine if the research proposals they are evaluating could be ‘reasonably anticipated to create, transfer, or use’ an ePPP? What guidance, if any, is provided to NIH program officials and leadership about how to evaluate whether proposed research could be ‘reasonably anticipated to create, transfer, or use’ an ePPP? How does NIAID determine which program officials and members of the institute’s leadership are appropriate reviewers of such proposals? Why are officials with expertise only in infectious diseases used in the review process and not a multidisciplinary approach that includes individuals with expertise in biosafety and biosecurity? If an NIAID official with infectious disease expertise determines that the proposed research may involve an enhanced potential pandemic pathogen, why does NIAID require a second internal review instead of referring that proposal directly to the HHS P3CO Review Group? How is the composition of this second-level review within NIAID determined and what criteria, standards, and process are used by this group to ‘reasonably anticipate’ if a research proposal would result in the creation, transfer, or use of an ePPP? What does it mean that high-level internal review at NIAID considers the research proposal in ‘the context of the state of the science in that field’? Such a criterion is not part of the HHS P3CO Framework. What official or office is responsible for overseeing implementation of the HHS P3CO within NIH and each institute and generating and maintaining relevant documentation?

The lack of guidance provided by NIH on how it implements the HHS P3CO Framework is in contrast with the detailed advice it has provided to government agencies and research institutions that conduct oversight of dual-use research of concern. Following the issuance of new policies on dual-use research oversight in 2012 and 2014, NIH published guidance, including tools⁵⁷ and case studies,⁵⁸ to assist with implementation of these policies. While intended primarily for institutions that receive Federal funding to conduct dual-use research of concern, the guidance noted that Federal agencies that conduct such research could also benefit from using this guidance to implement their own responsibilities for oversight of dual-use research. The NIH should take its own advice and provide clear guidance to potential grant applicants and the non-government scientists and NIH officials involved in the scientific peer review of grant proposals about the criteria for evaluating whether research proposals fall under the HHS P3CO Framework and delineate the process by which NIH will evaluate such proposals and make the necessary determinations as required under the HHS P3CO Framework.

Lack of transparency

NIH has also failed to provide a clear and complete account of how it has implemented the 2014 ‘gain of function’ funding pause and the HHS P3CO Framework. The Washington Post identified 18 projects funded by NIH between 2012 to 2020 that appear to include experiments that would enhance the virulence and/or the transmissibility of a pathogen, including eight that were approved for funding after 2017.⁵⁹ A separate review conducted by Mark Lipsitch identified five funded research projects that he believes should have been subject to review under the P3CO Framework.⁶⁰ NIH, however, only forwarded three research proposals for review by the HHS P3CO Review Group since adoption of the HHS P3CO Framework.⁶¹ The controversy surrounding Boston University’s engineering of a chimeric coronavirus raises new questions about similar NIH-funded projects that were not reviewed under the P3CO Framework. It would be advisable for the HHS P3CO Review Group to review other research proposals which were flagged during the scientific peer review process as potentially involving research with enhanced potential pandemic pathogens but were determined by NIH not to fall under the scope of the HHS P3CO Framework.

The way in which the public has learned about NIH’s implementation of its oversight responsibilities under the 2014 funding pause and the P3CO Framework does not inspire confidence in the NIH’s commitment to transparency. The most useful information has been provided by NIH primarily in response to Congressional inquiries, FOIA requests, and lawsuits. Without a better understanding of the process and criteria used by NIH to judge whether proposed research can be ‘reasonably anticipated’ to create a potential pandemic pathogen with enhanced virulence or transmissibility, and how it has applied these principles in practice, it will be difficult to both restore public confidence in this oversight and effectively strengthen such oversight in the future.

Recommendations

Given the grave biosafety and biosecurity implications of research with enhanced potential pandemic pathogens and gaps and weaknesses in the current oversight system for this research, we offer a set of recommendations that incorporate, but also go beyond, those offered by the NSABB.

In the short-term, the Biden Administration could use its executive authority to replace the HHS P3CO Framework with a single government-wide policy governing research with potential pandemic pathogens that would apply to all Federal agencies and institutions that receive funding from a Federal agency to conduct life sciences research, even if the work with an ePPP was not government-funded. Such an approach would bring the P3CO policy into alignment with the broader policy on dual-use research oversight adopted in 2012. Under that policy, all institutions that receive Federal funding to conduct or sponsor life research must review certain types of experiments for dual-use concerns ‘even if the research is not supported by USG funds.’⁶² This new policy would need to be accompanied by clear guidance to both Federal agencies and life sciences research institutions about how to identify research that falls within the scope of the policy, the process for reviewing the risks and benefits of such research, and criteria for determining the appropriate measures needed to conduct such research safely, securely, and responsibly. These recommendations are in line with those proposed by the NSABB in its 2023 proposal.⁶³

In the medium-term, the United States should adopt a comprehensive, national oversight system for laboratory biosafety, field biosafety, laboratory biosecurity, and dual-use research. This biorisk management system would be based on new legislation and cover both publicly and privately funded research in the life sciences. Although including privately funded research in the P3CO framework is critical, there are currently no mechanisms for its inclusion, unless the experiments are conducted with Select Agents.⁶⁴ Closing this critical gap would require either new legislation or amendment of the 2001 U.S.A. PATRIOT Act and the Public Health Security and Bioterrorism Preparedness and Response Act of 2002 which provide the legal basis for regulating the possession, use, and transfer of Select Agents and toxins by individuals and institutions. Implementing such a law would probably require the establishment of an independent biorisk management agency dedicated to overseeing biosafety, biosecurity, and dual-use research across the country.⁶⁵

Implementing the recommendations of this article and those proposed by NSABB will also require a comprehensive analysis of potential pandemic pathogen risks to better define the scope of an expanded government-wide policy for oversight of research with potential pandemic pathogens. One of the first activities of the new national biorisk management agency we propose is to systematically define the characteristics of pathogens that qualify them as potential pandemic pathogens and the types of experiments that could lead to the enhancement of their transmissibility and/or virulence.

The resulting risk assessment guidance would be useful for scientists considering undertaking research involving potential pandemic pathogens as well as researchers engaged in the peer-review process for grant proposals and journal manuscripts. This guidance would also be useful for officials in Federal agencies that conduct, sponsor, or fund life sciences research to determine if any such research is subject to review under the new government-wide oversight policy for potential pandemic pathogens.

To accomplish this goal, a multi-step process should be followed. First, researchers and the policy community must define the combination of pathogen characteristics that are of special concern, without resorting to a simple taxonomical list which would undermine a key strength of the framework. Suggestions for biosecurity and biosafety regimes that use phenotypes instead of taxonomy as the basis for risk assessment have been suggested.⁶⁶ Because the P3CO framework currently concerns itself with two properties to define risk, transmissibility and virulence, the combination of transmissibility and virulence that necessitates additional oversight must be clearly defined. Instead of discussing these characteristics in the abstract, real world and notional examples of potential pandemic pathogens should be used to establish these thresholds.⁶⁷ These examples should include a range of modified pathogens that would occupy the entire risk landscape as defined from most to least transmissible and most to least virulent. These examples could include a H1N1 1918 pandemic influenza virus with enhanced virulence, Yersinia pestis that is more transmissible and resistant to multiple antibiotics, Ebola hemorrhagic fever virus with enhanced transmissibility, SARS-CoV-1 virus that is as transmissible as the original SARS-CoV-2 strain, SARS-CoV-2 that is as virulent as SARS-CoV-1, a Hendra virus that is more transmissible, measles virus that evades protective vaccination, a H3N2 1968 that is as virulent as H1N1 1918 influenza, and a PR8 influenza virus that is as virulent as seasonal influenza.

By considering these types of examples as a group, a clear boundary should become apparent regarding the types of pathogens that exist and those that may be created via research that need enhanced oversight and controls as part of the P3CO framework. This boundary is a critical step in the improvement of the P3CO framework to clearly define the risk space the P3CO framework is concerned with and, by extension and equally importantly, the pathogens (and their modifications) that are clearly outside the scope of the P3CO framework. Without agreement on this boundary, the P3CO framework cannot accommodate the growth and expansion of pathogen research and would not provide the clarity that researchers need when planning experiments. This exact same approach could be extended to improve the DURC and Select Agent Regulations to flexibly accommodate modified pathogens outside their current limited lists (and perhaps could also be used to remove agents from oversight).

After this boundary is defined, the community must decide how certain a specific result from the experiment must be to fall within the scope of the P3CO framework: experiments that feasibly could result in enhancements to transmissibility and/or virulence or merely experiments that probably would result in enhancements to transmissibility and/or virulence. This step alone would inform the interpretation of the term ‘reasonably anticipated’ in the current guidance. To inform this deliberation, we suggest the community must use real world examples.

For example, consider if a new member of a virus family that includes some PPPs was found in fruit bats, but has yet to cause any known human infections. Given the complexity of host restriction, transmissibility, and pathogenicity in viruses, a research proposal to alter virus surface proteins to better bind to human receptors in vitro probably would not create a pathogen that is transmissible and pathogenic in humans. Conversely, because this experiment could adapt the novel virus to growth in human cells, it could feasibly create an enhanced PPP. Therefore, this experiment would be included in the scope of P3CO should the community adopt the more inclusive ‘feasibly could’ standard versus the stricter ‘probably would’ standard. Other experiments, described from the point of view of the research proposal stage (not the results), could be included to further solidify the scope of P3CO research.

Once the boundary and certainty criteria are established, examples should be provided of experiments on either side of the line, all of which are described from the perspective of proposed research (instead of assuming the outcome). The examples should describe specific experimental approaches, strains, and assays because all of these are pertinent to determine if an enhanced PPP may be created and, if created, if it would be identified as such. For example, consider a research proposal to create a large panel of mutants of seasonal H1N1 influenza virus by introducing pairwise amino acid substitutions observed in the pandemic 1918 H1N1 influenza strain. While such small changes individually are extremely unlikely to generate a strain of seasonal influenza that is as transmissible or pathogenic as the pandemic 1918 strain, the generation of a large library of mutants increases the chance that one would be created. This experiment would lead to a publishable result if enhanced pathogenicity were observed in at least one strain. The inclusion of the measurement of virulence in animals implies that should a more virulent mutant be generated, it would be identified. In comparison, a proposal that performs only cellular studies and does not assess for pathogenicity would not identify a strain with enhanced pandemic potential and therefore would not generate an information risk. If such an experiment generated a strain with enhanced properties that would result in a small increased biosafety risk, but the strain is stored in a freezer amongst a large panel of other mutants or discarded after the experiment is completed, the biosafety and biosecurity risks would be reduced although not eliminated.

This work would define, for pathogens that start outside of the established boundary of the P3CO framework, what types of experiments would fall within scope. For those pathogens that start far from the boundary, the degree of expected change in transmissibility and/or virulence would need to be greater and the certainty that such an increase would occur must be higher. For those pathogens that are near the boundary or already inside it, perhaps any feasible increase in transmissibility or virulence may be sufficient for the experiment to be subject to enhanced oversight. Conversely, for those pathogens that sit far from the boundary, greater certainty of significant increases in transmissibility or virulence may be required for enhanced oversight.

Reforming oversight of dual-use research, including experiments involving enhancement of potential pandemic pathogens, has been opposed on the grounds of the logistical burden this may impose on researchers, the fragmented nature of dual-use research oversight, the ambiguity of terms such as ‘reasonably anticipated,’ and the lack of guidance and educational and training resources for researchers.⁶⁸ These concerns are legitimate which is why our recommended reform of the P3CO Framework includes a rigorous process to better define the types of experiments that should be subject to stricter oversight. In addition, the creation of an independent biorisk management agency would create an entity with the mission of providing guidance, education, training, and other support to scientists and research institutions to facilitate oversight of dual-use research and ensure that such research is being conducted safely, securely, and responsibly. Although this expansion of biorisk management authority in the United States may seem daunting, Canada has successfully implemented a similar model with their Human Pathogens and Toxins Act providing a legal foundation and the Centre for Biosecurity in the Public Health Agency of Canada responsible for implementation.⁶⁹ Such a comprehensive, national system is consistent with the goal of the Biden Administration to, ‘Ensure all facilities in the United States or funded by the U.S. Government that conduct life sciences research, development, manufacturing, or diagnostic activities with especially dangerous biological materials – or which conduct work reasonably anticipated to result in such materials – are implementing and maintaining effective, transparent, rigorous, and comprehensive oversight, training, and monitoring programs for biosafety, biosecurity, and responsible and ethical conduct in science.’⁷⁰

Conclusion

COVID-19 has reinforced the importance of understanding viral threats to global health security and revealed new vulnerabilities in our pandemic preparedness and response capabilities. Ironically, greater efforts to prevent future pandemics and to strengthen biopreparedness could lead to increased risks of accidental or deliberate pandemics occurring. Meanwhile, biorisk management policies in the United States, and elsewhere, are failing to keep up with these new threats.⁷¹ This article has provided an assessment of the strengths and weaknesses of the HHS P3CO Framework and recommendations for strengthening oversight of research with potential pandemic pathogens based on five years of experience with this policy. Given the growing interest in studying potential pandemic pathogens, more robust oversight of this class of research is vitally necessary. Our recommendations for how to achieve this outcome are consistent with those proposed by NSABB and our proposal for a comprehensive analysis of the risks posed by potential pandemic pathogen could be used to inform implementation of NSABB’s proposed oversight system. In the longer term, oversight of research with potential pandemic pathogens needs to be integrated into a more comprehensive national system for biorisk management that ensures that life sciences research in the United States is being conducted safely, securely, and responsibly.

Funding Statement

The author(s) reported there is no funding associated with the work featured in this article.

Disclosure statement

No potential conflict of interest was reported by the author(s).